CN107025747A - The smart home intruding detection system and detection method of space-time two-dimensional information fusion - Google Patents
The smart home intruding detection system and detection method of space-time two-dimensional information fusion Download PDFInfo
- Publication number
- CN107025747A CN107025747A CN201710396810.7A CN201710396810A CN107025747A CN 107025747 A CN107025747 A CN 107025747A CN 201710396810 A CN201710396810 A CN 201710396810A CN 107025747 A CN107025747 A CN 107025747A
- Authority
- CN
- China
- Prior art keywords
- evidence
- time
- sensor
- fusion
- smart home
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
- G08B29/18—Prevention or correction of operating errors
- G08B29/183—Single detectors using dual technologies
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Geophysics And Detection Of Objects (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses a kind of intrusion detection method of the smart home intruding detection system of space-time two-dimensional information fusion, it is characterized in that, comprise the following steps：1）Obtain evidence collectionE1；2）Obtain evidence collection E2；3）Fusion；4）Judge；5）Processing：Such as fusion results are less than threshold value, then return to step 1）；Such as fusion results are more than or equal to threshold value, then information process unit sends invasion prompt message to mobile receiving unit and alarmed, and this method practicality is good, can improve the accuracy rate and real-time of intrusion detection.The present invention also discloses a kind of smart home intruding detection system of space-time two-dimensional information fusion, this system group network is convenient, cost is low.
Description
Technical field
The present invention relates to intelligent security guard detection technique field, more particularly to a kind of smart home of space-time two-dimensional information fusion
Intruding detection system and detection method.
Background technology
Smart home is also known as intelligent dwelling, refers to using advanced computer, network service, the technology such as automatically controls,
The various application subsystems relevant with family life are organically combined, by integrated management, allow family life it is more comfortable,
Safely, effectively and energy-conservation, wherein intelligent household security system has very important status as subsystem therein.Intelligent family
It is that the multiple technologies such as sensing technology, radiotechnics, fuzzy control technology and information fusion technology are integrated to occupy safety-protection system
Integrated application, security situation in man can in real time be monitored for detecting and preventing external illegal invasion, invasion warn
Feelings inform user in time, and are alarmed by network.
Although intelligent household security system drastically increases our quality of the life, makes our life become more to relax
Accommodate and feel at ease.But the system still has some problems in actual applications, walking about for such as indoor pet causes intrusion detection
System is reported by mistake, and invasion situation about failing to report is there is also under individual cases, this mainly due to the intrusion detection algorithm used for
Single-sensor threshold value diagnostic method, its false alarm and false dismissal probability are higher.Introduce after intelligent Intrusion Detection Technique, be based on
The information fusion detection algorithm of Dempster-Shafer (abbreviation D-S) evidence theory is main detection mode.Multisensor D-S
Fusion rule realizes the Intelligent Fusion of multi-sensor information, improves accuracy of detection, but D-S information fusions detection algorithm without
Method directly merges the sensing data of the high and complete contradiction of conflict degree, when sensor because the data that failure or interference acquisition are arrived
When differing greatly, blending algorithm can obtain the result of decision runed counter to the fact, because the sensing data of contradiction result in
The phenomenon of " veto by one vote " in D-S algorithms and rationally correct decision-making can not be made；In addition, the result of D-S fusions needs to adopt
The result that decision-making goes out detection can be just restrained in the case of collecting a large amount of evidences, if convergence rate will be directly affected very much slowly
To the real-time of detecting system, this causes serious influence for the practical significance that intelligent safety and defence system is present.
The content of the invention
The purpose of the present invention is in view of the shortcomings of the prior art, and to provide a kind of smart home of space-time two-dimensional information fusion
Intruding detection system and detection method.
This system group network is convenient, cost is low.
This method practicality is good, can improve the accuracy rate and real-time of intrusion detection.
Realizing the technical scheme of the object of the invention is：
A kind of smart home intruding detection system of space-time two-dimensional information fusion, including information process unit and with information
The sensor network and mobile receiving unit of unit connection are managed, the sensor network is provided with least two sensor unit.
Described information processing unit includes processor and the memory, the first radio receiving transmitting module and the net that are connected with processor
Network transceiver module.
The sensor unit includes sensor assembly, D/A converter module, the processor module and second being linked in sequence
Radio receiving transmitting module, the second radio receiving transmitting module is connected with the first radio receiving transmitting module.
The mobile receiving unit is provided with mobile terminal, and mobile terminal is connected with network transceiving module.
The sensor assembly is infrared, sound, vibrations, microwave remote sensor.
Sensor network, in real time gather domestic environment physical message, then by the second radio receiving transmitting module by this
A little information are sent to the first radio receiving transmitting module on information process unit；
Information process unit is used for the data message that real-time reception sensor unit is sent, and carries out data fusion, examines
Survey and whether invaded in family；If any, then send prompting message to move receiving unit.
Using the intrusion detection method of the smart home intruding detection system of above-mentioned space-time two-dimensional information fusion, including it is as follows
Step：
1) evidence collection E1 is obtained：Using the data of all node actual measurements in sensor network as evidence, according to Time Domain Fusion
It is regular to be handled, be specially：The measured data that any sensor i in household safety-protection network is obtained in t is according to probability
Mapping ruler function mi,t,p() is proposition set Θ={ someone, pet, no one } allocation probability initial value, is then made the biography
The real-time evidence E of sensori,t,p={ mi,t,p(someone), mi,t,p(pet), mi,t,p(no one) }, next according to time-domain adaptive
Weighted Fusion rule is merged to the evidence, obtains the accumulation of evidence E of the current time sensori,t,c={ mi,t,c(have
People), mi,t,c(pet), mi,t,c(no one) }, the accumulation of evidence of all the sensors is combined and then evidence collection E1 is obtained；
2) evidence collection E2 is obtained：Evidence collection E1 is handled according to spatial domain fusion rule, i.e., by infrared, sound, vibrations,
Colliding datas of the accumulation of evidence collection E1 of microwave remote sensor composition in spatial domain is modified, and obtains revised evidence collection E2；
3) merge：Evidence collection E2 is merged according to the rule of information fusion, fusion results are obtained, i.e., by evidence collection E2
In evidence merged successively according to D-S evidences, the fusion for obtaining proposition set " someone, pet, no one " is general
Rate apportioning cost；
4) judge：By step 3) fusion results are compared with given threshold 0.7, determine whether invader, Ji Jiangbu
Proposition is compared for the fusion probable value of " someone " with given threshold 0.7 in rapid 3) fusion results, determines whether invader；
5) handle：Such as fusion results are less than threshold value, then return to step 1)；Such as fusion results are more than or equal to threshold value, then believe
Breath processing unit sends invasion prompt message to mobile receiving unit and alarmed.
Step 1) described in Time Domain Fusion rule specifically include：
To any any proposition cumulative probability value computational charts of the single sensor i included in the accumulative evidence of t
It is up to formula
mi,t,c(A)=αi,t-1,c,i,t,pmi,t-1,c(A)+βi,t-1,c,i,t,pmi,t,p(A),
Wherein, it is any proposition in proposition set, m that A, which is,i,t,c(A) to be right in the current t cumulative evidence of the sensor
Answer proposition A cumulative probability values, mi,t-1,c(A) the cumulative probability value for being correspondence proposition A in sensor t-1 hours cumulative evidences；
mi,t,p(A) the real-time probable value for being proposition A in the sensor t real-time evidence；αi,t-1,c,i,t,pIt is the sensor in t
Real-time evidence and the cumulative evidence at t-1 moment between coefficient of similarity, scope be 0 to 1；βi,t-1,c,i,t,cFor conflict coefficient,
βi,t-1,c,i,t,p=1- αi,t-1,c,i,t,p.Coefficient of similarity α between evidencei,t-1,c,i,t,pCalculation be：Wherein ki,t-1,c,i,t,pFor t evidence Ei,t,pWith t-1 moment accumulation of evidence
Ei,t-1,cD-S evidence theory conflict value,For t evidence Ei,t,pWith t-1 moment accumulation of evidence Ei,t-1,c's
Pignistic probability metricses.
Step 2) described in spatial domain fusion rule comprise the following steps：
2.1)：Each sensor cumulative evidence in evidence collection E2 arbitrarily sorts, then by their end to end compositions one
Individual annular evidence sequence, calculates the coefficient of similarity between the adjacent evidence of any two；
2.2)：Strong correlation threshold value and weak dependence threshold value are designated as η respectively between setting evidenceERAnd ηIRIf, annular evidence sequence
A certain bar evidence in row is all higher than η with the coefficient of similarity of its previous bar evidence and the coefficient of similarity of latter bar evidenceER, then
It is strong correlation evidence to mark the evidence；If respectively less than ηIR, then the evidence weak dependence evidence is marked；Remaining evidence is then marked
For general correlation evidence；
2.3)：By the discarding of weak relevant evidence, the reservation of strong correlation evidence, general correlation evidence with evidence collection desired value to it
Replace amendment.
Step 2.3) described in evidence collection it is desired calculating comprise the following steps：
2.3.1)：The support that single evidence is concentrated in evidence is calculated according to the coefficient of similarity between sensor evidence
Degree, i.e., according to single sensor i t cumulative evidence Ei,t,c={ mi,t,c(someone), mi,t,c(pet), mi,t,c(do not have
People) }, calculate its support in evidence collection E2Wherein
ki,t,c,j,t,cWithIt is evidence E respectivelyi,t,cWith evidence Ej,t,cThe theoretical conflict coefficients and Pignistic probability of D-S
Distance；
2.3.2)：Evidence concentrates the support of the single evidence institute than on support and to be worth to the power of the evidence on evidence
Value, the expectation of evidence collection is calculated by way of weighted sum, i.e., evidence is concentrated any one evidence Ei,t,cSupport Sup
(Ei,t,c) institute support and value on evidence on ratioObtain the weights of the evidencePass through weighting
Summing mode calculates evidence collection expectationWherein
The smart home security intruding detection system and detection method of this space-time two-dimensional information fusion can eliminate sensing
The problem of contradiction is higher between device gathered data, and then avoid D-S algorithms from producing " veto by one vote " phenomenon in decision-making and carry
The high convergent speed of fusion, the accuracy and the real-time of system that algorithm decision-making is improved with this, and then reduce smart home peace
The false alarm rate and false dismissed rate of anti-system.
This system group network is convenient, cost is low.
This method practicality is good, can improve the accuracy rate and real-time of intrusion detection.
Brief description of the drawings
Fig. 1 is the structural representation of system in embodiment；
Fig. 2 is method flow schematic diagram in embodiment；
Fig. 3 is the hollow time-domain information fusion principle schematic of embodiment；
Fig. 4 is intrusion detection fusion results schematic diagram in embodiment.
In figure, the modulus of 11. sensor assembly of the information process unit 3. of 1. sensor unit 2. movement receiving unit 12.
The radio receiving transmitting module 21. of 13. processor module of modular converter 14 second includes the processor module 23. first of memory 22.
The intelligent terminal of 24. network transceiving module of radio receiving transmitting module 31..
Embodiment
Present invention is further elaborated with reference to the accompanying drawings and examples, but is not limitation of the invention.
Embodiment：
Reference picture 1, a kind of smart home intruding detection system of space-time two-dimensional information fusion, including information process unit 2
With the sensor network being connected with information process unit 2 and mobile receiving unit 3, the sensor network is passed provided with least two
Sensor cell 1.
Described information processing unit 2 includes processor 22 and the memory 21, the first wireless receiving and dispatching that are connected with processor 22
Module 23 and network transceiving module 24.
The sensor unit 1 includes sensor assembly 11, D/A converter module 12, the processor module 13 being linked in sequence
With the second radio receiving transmitting module 14, the second radio receiving transmitting module 14 is connected with the first radio receiving transmitting module 23.
The mobile receiving unit 3 is provided with mobile terminal 31, and mobile terminal 31 is connected with network transceiving module 24.
The sensor assembly 11 is infrared, sound, vibrations, microwave remote sensor.
Sensor network, then will by the second radio receiving transmitting module 14 for gathering domestic environment physical message in real time
These information are sent to the first radio receiving transmitting module 23 on information process unit 2；
Information process unit 2 is used for the data message that sends of real-time reception sensor unit 1, and carries out data fusion,
Whether invaded in detection man；If any, then send prompting message to move receiving unit 31.
In this example, the structure of sensor unit 1 is the pin of sensor and is equipped with the CC2530 cores of Zigbee protocol
The I/O mouths of piece are attached, and realize the collection of household physical context information.
In this example, the structure of information process unit 2 is serial ports and Ti companies the CC2530 development boards of Ti companies
The serial ports of AM3358 development boards is attached, and is realized the data that receiving sensing unit 1 is transmitted, is connect by Ethernet network interface to movement
Receive the transmission data of mobile terminal 31 of unit 3.
Reference picture 2, using the intrusion detection method of the smart home intruding detection system of above-mentioned space-time two-dimensional information fusion,
Comprise the following steps：
1) evidence collection E1 is obtained：Using the data of all node actual measurements in sensor network as evidence, according to Time Domain Fusion
It is regular to be handled, be specially：The measured data that any sensor i in household safety-protection network is obtained in t is according to probability
Mapping ruler function mi,t,p() is proposition set Θ={ someone, pet, no one } allocation probability initial value, is then made the biography
The real-time evidence E of sensori,t,p={ mi,t,p(someone), mi,t,p(pet), mi,t,p(no one) }, next according to time-domain adaptive
Weighted Fusion rule is merged to the evidence, obtains the accumulation of evidence E of the current time sensori,t,c={ mi,t,c(have
People), mi,t,c(pet), mi,t,c(no one) }, the accumulation of evidence of all the sensors is combined and then evidence collection E1 is obtained；
2) evidence collection E2 is obtained：Evidence collection E1 is handled according to spatial domain fusion rule, i.e., by infrared, sound, vibrations,
Colliding datas of the accumulation of evidence collection E1 of microwave remote sensor composition in spatial domain is modified, and obtains revised evidence collection E2；
3) merge：Evidence collection E2 is merged according to the rule of information fusion, fusion results are obtained, i.e., by evidence collection E2
In evidence merged successively according to D-S evidences, the fusion for obtaining proposition set " someone, pet, no one " is general
Rate apportioning cost；
4) judge：By step 3) in fusion results proposition compared for the fusion probable value of " someone " with given threshold 0.7
Compared with determining whether invader；
5) handle：Such as fusion results are less than threshold value, then return to step 1)；Such as fusion results are more than or equal to threshold value, then believe
Breath processing unit 2 sends invasion prompt message to mobile receiving unit 3 and alarmed.
Specifically, this example step 1) in, described mapping ruler function mi,t,p() can be according to such as lower probability mapping table
It is configured：
During classical D-S is theoretical, two evidence E1And E2Between conflict coefficient be expressed as：
Wherein A and B are the proposition event in proposition set Θ, it is clear that k1,2Closer to 1, show between two evidences
Conflict it is bigger, however, the conflict spectrum that such a mode is characterized between evidence has open defect, such as two are identical
Evidence E1={ 0.7,0.3 } and E2={ 0.7,0.3 }, the conflict coefficient between them should be 0, however, the conflict system that above formula is calculated
Number is k1,2=0.42；
In order to improve this defect, introduce Pignistic probability metricses to correct the expression of conflict coefficient, it is assumed that E1With
E22 evidences collected for sensor network node, then E1And E2Between Pignistic probability metrics expression formulas be：
Wherein,For evidence EiPignistic probability letters just on proposition complete or collected works
Number, the E after Pignistic probability metrics amendments1And E2Between conflict coefficient be represented byAccordingly
Coefficient of similarity be represented by α1,2=1- β1,2。
The conventional method of Time Domain Fusion is D-S fusions or Weighted Fusion, but the frequency extremely conflicted occurs in time domain data
Very high, Weighted Fusion is due to being manually set weights so that fusion results are unreasonable, therefore weighted in this example using time-domain adaptive
Fusion rule, it is to avoid the conflict spectrum height between evidence, which brings the fusion results undesirable and is manually set the methods of weights, not to be conformed to
Reason, its implementation process is step 1), step is implemented referring to Fig. 3.
Specifically, this example step 1) in, Time Domain Fusion rule includes：
To any sensor i in infrared, sound, vibrations, microwave remote sensor in t cumulative evidence Ei,t,cIncluded in
Proposition A cumulative probability value calculation expression is：
mi,t,c(A)=αi,t-1,c,i,t,pmi,t-1,c(A)+βi,t-1,c,i,t,pmi,t,p(A)；
Wherein, mi,t,c(A) be infrared, sound, vibrations, microwave remote sensor in any sensor i in t cumulative evidence
Ei,t,cMiddle proposition A cumulative probability value, mi,t-1,c(A) for the sensor in t-1 hours cumulative evidences Ei,t-1,cMiddle proposition A's is tired
Product probable value, mi,t,p(A) for the sensor in t real-time evidence Ei,t,pMiddle proposition A real-time probable value, ai,t-1,c,i,t,p=
1-βi,t-1,c,i,t,pFor the sensor t real-time evidence Ei,t,pWith the cumulative evidence E at t-1 momenti,t-1,cBetween it is similar
Property coefficient.Wherein,For conflict coefficient.
Wherein, ki,t-1,c,i,t,pFor the sensor t real-time evidence Ei,t,pWith the cumulative evidence at t-1 moment
Ei,t-1,cBetween D-S evidence theory conflict value,Represent real-time evidence E of the sensor in ti,t,pWith t-1
The cumulative evidence E at momenti,t-1,cBetween Pignistic probability metricses.For example, the card that infrared sensor is collected in t
According to Er,t,pFor mr,t,p(someone), mr,t,p(pet), mr,t,p(no one), its accumulation of evidence E at the t-1 momentr,t-1,cFor mr,t-1,c
(someone), mr,t-1,c(pet), mr,t-1,c(no one), then infrared sensor be in the expression formula of the cumulative evidence of t：
mr,t,c(someone)=αr,t,p,r,t-1,cmr,t-1,c(someone)+βr,t,p,r,t-1,cmr,t,p(someone)；
mr,t,c(pet)=αr,t,p,r,t-1,cmr,t-1,c(pet)+βr,t,p,r,t-1,cmr,t,p(pet)；
mr,t,c(no one)=αr,t,p,r,t-1,cmr,t-1,c(no one)+βr,t,p,r,t-1,cmr,t,p(no one).
Step 2) described in spatial domain fusion rule comprise the following steps：
2.1)：Each sensor cumulative evidence in evidence collection E2 arbitrarily sorts, then by their end to end compositions one
Individual annular evidence sequence, calculates the coefficient of similarity between the adjacent evidence of any two；
2.2)：Strong correlation threshold value and weak dependence threshold value are designated as η respectively between setting evidenceERAnd ηIRIf, annular evidence sequence
A certain bar evidence in row is all higher than η with the coefficient of similarity of its previous bar evidence and the coefficient of similarity of latter bar evidenceER, then
It is strong correlation evidence to mark the evidence；If respectively less than ηIR, then the evidence weak dependence evidence is marked；Remaining evidence is then marked
For general correlation evidence；
2.3)：By the discarding of weak relevant evidence, the reservation of strong correlation evidence, general correlation evidence with evidence collection desired value to it
Replace amendment.
Fusion conventional method in spatial domain is that multisensor is merged with D-S rules, but classics D-S blending algorithms can not be direct
The high data of conflict degree are handled, when the sensor of some in evidence set because the evidence that failure or interference are obtained is deposited with other evidences
In extreme conflict, fusion results just occur that " veto by one vote " phenomenon, i.e. fusion results level off to failure or interference sensor
Evidence, for such case, this example corrects the rule of colliding data fusion, the sensing collected to current time using spatial domain
Device accumulation data are handled, and find out conflicting evidence and it is corrected with evidence collection desired value, so not only accelerate D-S fusions
As a result convergence rate, also causes fusion results more rationally effectively, and then improve the accuracy and real-time of detection means,
Its implementation process is step 2), step is implemented referring to Fig. 3, including：
This example step 2) in, described spatial domain fusion rule comprises the following steps：
Step 2.1)：Assuming that the accumulation of infrared sensor, sound transducer, shock sensor, microwave remote sensor in t
Evidence is designated as E respectivelyr,t,c、Ev,t,c、Es,t,c、Ew,t,c, first by their end to end annular evidence sequences of composition one, then
According to the definition to evidence similarity, the similarity that can obtain between the sensor cumulative evidence is followed successively by αr,t,c,v,t,c、
αv,t,c,s,t,c、αs,t,c,w,t,c、αw,t,c,r,t,c；
Step 2.2)：It is assumed that the threshold value η of strong correlation evidenceER=0.7, the threshold value η of weak dependence evidenceIR=0.3, if
The coefficient of similarity and the coefficient of similarity of latter bar evidence of a certain bar evidence and its previous bar evidence in annular evidence sequence
It is all higher than ηER, then the evidence is strong correlation evidence, if respectively less than ηIR, then the evidence is weak relevant evidence；Remaining evidence
For general correlation evidence, such as sound transducer evidence Ev,t,c, the coefficient of similarity of itself and previous bar evidence is
αr,t,c,v,t,c, the coefficient of similarity of itself and latter bar evidence is αv,t,c,s,t,c；If αr,t,c,v,t,cAnd αv,t,c,s,t,cIt is all higher than
0.7, then sound transducer evidence Ev,t,cFor strong correlation evidence；
Step 2.3), by weak dependence evidence abandon, strong correlation evidence retain, general correlation evidence with evidence collection expect
Amendment is replaced to it.
Step 2.3) described in evidence collection it is desired calculating comprise the following steps：
2.3.1)：The support that single evidence is concentrated in evidence is calculated according to the coefficient of similarity between sensor evidence
Degree, i.e., according to single sensor i t cumulative evidence Ei,t,c={ mi,t,c(someone), mi,t,c(pet), mi,t,c(do not have
People) }, calculate its support in evidence collection E2Wherein
ki,t,c,j,t,cWithIt is evidence E respectivelyi,t,cWith evidence Ej,t,cThe theoretical conflict coefficients and Pignistic probability of D-S
Distance；
2.3.2)：Evidence concentrates the support of the single evidence institute than on support and to be worth to the power of the evidence on evidence
Value, the expectation of evidence collection is calculated by way of weighted sum, i.e., evidence is concentrated any one evidence Ei,t,cSupport Sup
(Ei,t,c) institute support and value on evidence on ratioObtain the weights of the evidenceAsked by weighting
The evidence collection is calculated with mode to expectWherein
Specifically, this example step 2.3) in, the desired calculation procedure of evidence collection is as follows：
The cumulative evidence point of known infrared sensor, sound transducer, shock sensor, microwave remote sensor in same t
E is not designated as itr,t,c、Ev,t,c、Es,t,c、Ew,t,c；
Step 2.3.1)：Defined according to coefficient of similarity between evidence, calculate the similarity system between any two sensor
Number αr,t,c,v,t,c、αr,t,c,s,t,c、αr,t,c,w,t,c、αv,t,c,s,t,c、αv,t,c,w,t,c、αs,t,c,w,t,c；Then infrared, sound is calculated
The support of the respective sensors such as sound, vibrations, microwave is respectively
Sup(Er,t,c)=αr,t,c,v,t,c+αr,t,c,s,t,c+αr,t,c,w,t,c
Sup(Ev,t,c)=αr,t,c,v,t,c+αv,t,c,s,t,c+αv,t,c,w,t,c
Sup(Es,t,c)=αr,t,c,s,t,c+αv,t,c,s,t,c+αs,t,c,w,t,c
Sup(Ew,t,c)=αr,t,c,w,t,c+αv,t,c,w,t,c+αs,t,c,w,t,c；
Step 2.3.2)：With the support of single evidence than on support and be worth to the weights of the evidence on evidence, with
This mode, which is calculated, to be obtained the weights of the sensors such as infrared, sound, vibrations, microwave and is respectively Wherein Ω={ r, v, s, w }；
Evidence collection is calculated in weighted sum mode to be desired forWherein
Specifically, step 3 in this example) in, the evidence in evidence set E2 is melted according to D-S fusion rules successively
Close, specially fusion rule includes：
For any two evidence E in evidence set E2i={ mi(someone), mi(pet), mi(no one) } and Ej={ mj
(someone), mj(pet), mj(no one) }, D-S fusion rules are
Wherein,For D-S conflict of theories values.
Specifically, step 4 in this example) in, by step 3) in obtain proposition in result " someone " according to D-S fusion rules
Fusion probable value is compared in 0.7, is specially：
For step 3) obtained fusion results Eresult={ mresult(someone), mresult(pet), mresult(no one) },
By mresult(someone) is compared with decision-making value 0.7, and decision-making is no to there is intrusion event if more than or equal to 0.7
Then, decision-making is to occur without intrusion event.
In this example, experimental simulation invader collect data and its fusion results into various kinds of sensors as shown in figure 4,
3 stages of experiment point simulate the testing result in different invasion situations and working sensor state：
First stage (sampling order 0-15), the situation that simulation invader enters：It can be seen by Fig. 4 sampling order 0-15
Go out invader at the 3rd sampling period to enter, 4 class sensors have sensed someone's entrance, underlying probabilities apportioning cost also phase
It should increase, final fusion detection result converges on someone's entrance；
Second stage (sampling order 15-25), any of simulation detection system sensor break down independent or
By the data in the case of external interference：It can be seen that four class sensors respectively by dry by Fig. 4 sampling order 0-15
When disturbing, the data gathered are not impacted to final fusion results, and someone will not be detected because of being interfered
Presence；
In phase III (sampling order 25-45), simulation detection system any two sensors simultaneously break down or by
Data in the case of interference：Even if can be seen that the feelings being interfered in 2 sensors by Fig. 4 sampling order 25-45
Under condition, the result that fault data is obtained to final fusion is to cause to significantly affect；
Step 5)：Such as step 4) decision-making goes out is the generation for having intrusion behavior, then and information process unit 2 is to mobile receiving unit
3 send invasion prompt message and alarm.
Claims (8)
1. a kind of smart home intruding detection system of space-time two-dimensional information fusion, it is characterized in that, including information process unit and
The sensor network and mobile receiving unit being connected with information process unit, the sensor network are provided with least two sensor
Unit.
2. the smart home intruding detection system of space-time two-dimensional information fusion according to claim 1, it is characterized in that, it is described
Information process unit includes processor and the memory, the first radio receiving transmitting module and the network transceiving module that are connected with processor.
3. the smart home intruding detection system of space-time two-dimensional information fusion according to claim 1, it is characterized in that, it is described
Sensor unit includes sensor assembly, D/A converter module, processor module and the second radio receiving transmitting module being linked in sequence,
Second radio receiving transmitting module is connected with the first radio receiving transmitting module.
4. a kind of smart home intruding detection system of the space-time two-dimensional information fusion described in use claim any one of 1-3
Intrusion detection method, it is characterized in that, comprise the following steps：
1) evidence collection E1 is obtained：Using the data of all node actual measurements in sensor network as evidence, according to Time Domain Fusion rule
Handled, obtain evidence collection E1；
2) evidence collection E2 is obtained：Evidence collection E1 is handled according to spatial domain fusion rule, evidence collection E2 is obtained；
3) merge：Evidence collection E2 is merged according to the rule of information fusion, fusion results are obtained；
4) judge：By step 3) fusion results are compared with given threshold 0.7, determine whether invader；
5) handle：Such as fusion results are less than threshold value, then return to step 1)；Such as fusion results are more than or equal to threshold value, then at information
Reason unit sends invasion prompt message to mobile receiving unit and alarmed.
5. the smart home intrusion detection method of space-time two-dimensional information fusion according to claim 4, it is characterized in that, step
1) the Time Domain Fusion rule described in is specifically included：
To any any proposition cumulative probability value calculation expressions of the single sensor i included in the accumulative evidence of t
For
mi,t,c(A)=αi,t-1,c,i,t,pmi,t-1,c(A)+βi,t-1,c,i,t,pmi,t,p(A),
Wherein, it is any proposition in proposition set, m that A, which is,i,t,c(A) it is correspondence life in the current t cumulative evidence of the sensor
Inscribe A cumulative probability values, mi,t-1,c(A) the cumulative probability value for being correspondence proposition A in sensor t-1 hours cumulative evidences；mi,t,p
(A) the real-time probable value for being proposition A in the sensor t real-time evidence；αi,t-1,c,i,t,pFor the sensor t reality
When evidence and the cumulative evidence at t-1 moment between coefficient of similarity, scope be 0 to 1；βi,t-1,c,i,t,cFor conflict coefficient,
βi,t-1,c,i,t,p=1- αi,t-1,c,i,t,p。
6. the smart home intrusion detection method of space-time two-dimensional information fusion according to claim 5, it is characterized in that, it is described
Coefficient of similarity α between evidencei,t-1,c,i,t,pCalculation be：Wherein
ki,t-1,c,i,t,pFor t real-time evidence Ei,t,pWith t-1 moment accumulation of evidence Ei,t-1,cD-S evidence theory conflict value,For t real-time evidence Ei,t,pWith t-1 moment accumulation of evidence Ei,t-1,cPignistic probability metricses.
7. the smart home intrusion detection method of space-time two-dimensional information fusion according to claim 4, it is characterized in that, step
2) the spatial domain fusion rule described in comprises the following steps：
2.1)：Each sensor cumulative evidence in evidence collection E2 arbitrarily sorts, then by their one rings of end to end composition
Shape evidence sequence, calculates the coefficient of similarity between the adjacent evidence of any two；
2.2)：Strong correlation threshold value and weak dependence threshold value are designated as η respectively between setting evidenceERAnd ηIRIf, in annular evidence sequence
The coefficient of similarity of a certain bar evidence and its previous bar evidence and the coefficient of similarity of latter bar evidence be all higher than ηER, then mark
The evidence is strong correlation evidence；If respectively less than ηIR, then the evidence weak dependence evidence is marked；Remaining evidence is then labeled as one
As correlation evidence；
2.3)：Weak relevant evidence is abandoned, strong correlation evidence retains, general correlation evidence is replaced with evidence collection desired value to it
Amendment.
8. the smart home intrusion detection method of space-time two-dimensional information fusion according to claim 7, it is characterized in that,
Step 2.3) described in evidence collection it is desired calculating comprise the following steps：
2.3.1)：The support that single evidence is concentrated in evidence is calculated according to the coefficient of similarity between sensor evidence；
2.3.2)：Evidence concentrates the support of the single evidence institute than on support and to be worth to the weights of the evidence on evidence, logical
The mode for crossing weighted sum calculates the expectation of evidence collection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710396810.7A CN107025747B (en) | 2017-05-31 | 2017-05-31 | The smart home intruding detection system and detection method of space-time two-dimensional information fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710396810.7A CN107025747B (en) | 2017-05-31 | 2017-05-31 | The smart home intruding detection system and detection method of space-time two-dimensional information fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107025747A true CN107025747A (en) | 2017-08-08 |
| CN107025747B CN107025747B (en) | 2019-06-14 |
Family
ID=59529970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710396810.7A Active CN107025747B (en) | 2017-05-31 | 2017-05-31 | The smart home intruding detection system and detection method of space-time two-dimensional information fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107025747B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108428315A (en) * | 2018-03-05 | 2018-08-21 | 中国人民解放军军事科学院国防工程研究院 | Electronic sentry system |
| CN108769015A (en) * | 2018-05-29 | 2018-11-06 | 深圳汇通智能化科技有限公司 | A kind of smart home intruding detection system |
| CN111784986A (en) * | 2020-07-13 | 2020-10-16 | 广元量知汇科技有限公司 | Intelligent security alarm method based on big data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5471194A (en) * | 1993-03-23 | 1995-11-28 | Aritech Corporation | Event detection system with centralized signal processing and dynamically adjustable detection threshold |
| CN101556651A (en) * | 2009-04-15 | 2009-10-14 | 北京航空航天大学 | Multi-source data fusion method in clustering wireless sensor network |
| CN202394353U (en) * | 2012-01-05 | 2012-08-22 | 广州铁路职业技术学院 | Household safety protector based on wireless network |
| CN102737468A (en) * | 2012-07-11 | 2012-10-17 | 重庆大学 | System and method for detecting fire hazard based on wireless multi-sensor information fusion |
| CN102945590A (en) * | 2012-10-19 | 2013-02-27 | 安徽工程大学 | Home security system based on wireless sensing network |
| CN104616416A (en) * | 2015-01-14 | 2015-05-13 | 东华大学 | Multi-sensor information fusion-based wireless fire alarm system |
-
2017
- 2017-05-31 CN CN201710396810.7A patent/CN107025747B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5471194A (en) * | 1993-03-23 | 1995-11-28 | Aritech Corporation | Event detection system with centralized signal processing and dynamically adjustable detection threshold |
| CN101556651A (en) * | 2009-04-15 | 2009-10-14 | 北京航空航天大学 | Multi-source data fusion method in clustering wireless sensor network |
| CN202394353U (en) * | 2012-01-05 | 2012-08-22 | 广州铁路职业技术学院 | Household safety protector based on wireless network |
| CN102737468A (en) * | 2012-07-11 | 2012-10-17 | 重庆大学 | System and method for detecting fire hazard based on wireless multi-sensor information fusion |
| CN102945590A (en) * | 2012-10-19 | 2013-02-27 | 安徽工程大学 | Home security system based on wireless sensing network |
| CN104616416A (en) * | 2015-01-14 | 2015-05-13 | 东华大学 | Multi-sensor information fusion-based wireless fire alarm system |
Non-Patent Citations (1)
| Title |
|---|
| 罗大庸、张远: "多传感器信息时空融合模型及算法研究", 《系统工程与电子技术》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108428315A (en) * | 2018-03-05 | 2018-08-21 | 中国人民解放军军事科学院国防工程研究院 | Electronic sentry system |
| CN108769015A (en) * | 2018-05-29 | 2018-11-06 | 深圳汇通智能化科技有限公司 | A kind of smart home intruding detection system |
| CN111784986A (en) * | 2020-07-13 | 2020-10-16 | 广元量知汇科技有限公司 | Intelligent security alarm method based on big data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107025747B (en) | 2019-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107025747B (en) | The smart home intruding detection system and detection method of space-time two-dimensional information fusion | |
| Dong et al. | Combination of evidential sensor reports with distance function and belief entropy in fault diagnosis | |
| CN103840967B (en) | A kind of method of fault location in power telecom network | |
| JP5410446B2 (en) | System and method for object location and path identification based on RFID detection | |
| CN107241226A (en) | Fuzz testing method based on industry control proprietary protocol | |
| CN101355504A (en) | Method and apparatus for confirming user behavior | |
| CN100362803C (en) | Network safety warning system based on cluster and relavance | |
| CN101494535A (en) | Method for constructing network inbreak scene based on hidden Mrakov model | |
| CN109150868A (en) | network security situation evaluating method and device | |
| CN102255764A (en) | Method and device for diagnosing transmission network failure | |
| CN107463106B (en) | Intelligent home system | |
| CN106685742A (en) | Network fault diagnosing method and apparatus | |
| CN105337957A (en) | SDN network DDoS and DLDoS distributed space-time detection system | |
| CN106411904A (en) | Network risk control method based on microstate prediction | |
| CN106462547A (en) | Maintaining routing information | |
| CN109948695A (en) | A kind of power grid fragility node automatic identifying method based on neighbour's propagation clustering algorithm | |
| CN108882152B (en) | User privacy protection method based on path selection reporting | |
| CN107733877A (en) | A kind of management method and system of Internet of Things wireless telecommunications framework | |
| Tamani et al. | Rule-based model for smart building supervision and management | |
| CN102195975A (en) | Intelligent NIPS (Network Intrusion Prevention System) framework for quantifying neural network based on mobile agent (MA) and learning vector | |
| CN107623615A (en) | A kind of intelligent home device control system based on wechat platform | |
| CN104239785B (en) | Intrusion detection data classification method based on cloud model | |
| CN105959184A (en) | Smart indoor air purification system | |
| Rahmani et al. | Sensor selection for detecting deviations from a planned itinerary | |
| CN103246921B (en) | A kind of underwater robot environment perception method based on immune main body |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |